‘Soft’ and ‘Law’
The words composing the phrase provide a starting point for understanding some of the existing conceptions and usages of this term. Some definitions focus on the word ‘soft’ as giving the phrase distinctive meaning. According to such definitions, ‘soft’ law is a type of law—that is, a law that is in some sense ‘soft’.
This camp is then subdivided between those who adopt a binary approach between law and non-law, and those who view law as more of a continuum. Some adopt a binary approach, where something is either law or not law. For those adhering to this more positivist approach, ‘soft law’ refers to rules of positive law that are recognized as such, but nonetheless may be considered soft in their commands. This softness might be attributable to, for example, vague language, non-justiciable formulations, or an exclusive reliance on auto-interpretation by the subject of obligation. For those who adopt this approach, there is no such thing as a soft law instrument, for it is not the character of the instrument that makes it soft. In order for something to be called soft law, it must be existing international law according to the established sources.
Others within this camp reject the binary approach and rather see a broader spectrum of law than what is formulated in Article 38 of the ICJ Statute.  For this second group, soft law need not be law established in accordance with the traditional doctrine of sources. Soft law may thus not be legally binding in the technical sense, but the adherents of this view tend to accord the binding/non- binding dichotomy less significance, instead viewing a spectrum of normativity and behaviour-influencing capacity. They view all of it as a type of ‘law’ because the norms are typically formulated as rules and are intended to, and often do, guide behaviour. The ‘softness’ of a norm, formulation, or instrument could be in terms of, inter alia, specificity, onerousness, enforceability, justiciability, normativity, or ‘bindingness’.5
Rather than seeing soft law as a type of law, a third approach, which is also the approach of the present author, is to not regard ‘soft law’ as law at all. This conception also adopts a positivistic, binary perspective, but is the inverse of the positivistic perspective described above, in that it entirely excludes positive law from the ambit of soft law and includes only rules that do not have the status of existing international law according to the established sources. According to this view, the term ‘law’ in the phrase soft law does not describe the status of the rule itself, but instead refers to the form or provenance of the rule and may also reflect its relationship to and potential relevance in the international legal system. In this sense, the rule is described as ‘soft’ precisely because it is not itself law, but is still referred to as ‘law’ because it may exert an influence on law or may achieve some of the effects that law seeks to achieve.
All three of the above approaches are reflected in usage of the term ‘soft law’ by both academics and practitioners, though most common usages seem to correspond to the second and third conceptions.
Another variation in usage should be noted. Most of the views formulated above conceive of soft law as a category of norms; however, the term is also used to describe a category of instruments—generally, instruments that are adopted in an intergovernmental setting and that set forth rules of behaviour or other normative standards, but that are not legally binding. This usage would not correspond to the notion of soft law set forth in the first conception above, as that conception is limited to norms that formally have the status of positive law, but it would overlap with the second and third conceptions. However, as one use of the term refers to the norms themselves and the other refers to the type of instrument, the usages are not coextensive. As such, the terms ‘soft law’ and ‘soft law instrument’ will each be defined separately in sections 3 and 4.
-  P. Weil, ‘Towards Relative Normativity in International Law’, American Journal of InternationalLaw vol. 77 (1983): 413-42 at 414.
-  See e.g.: D. Shelton (ed.), Commitment and Compliance: The Role of Non-binding Norms in theInternational Legal System (Oxford: Oxford University Press, 2000).
-  See e.g.: G. Shaffer and M. Pollack, ‘Hard Versus Soft Law in International Security’, BostonCollege Law Review vol. 52 (2011): 1147-241 at 1160; C. Christine, ‘The Challenge of SoftLaw: Development and Change in International Law’, International and Comparative Law Quarterlyvol. 38 (1989): 850-66 at 851.
-  See e.g.: Klabbers (1996): 167-82.
-  J. d’Aspremont, ‘The Politics of Deformalization in International Law’, Gottingen Journal ofInternational Law vol. 3 (2011): 503—50 at 526—7; Klabbers (1996): 168.